AI makes humans better at spotting cyber risks
Ethical human hackers supported by machine learning and artificial intelligence are 73 percent more efficient at identifying and evaluating cyber risks and threats according to a new report.
The study from crowdsourced security platform Synack also finds this combination of cybersecurity talent and AI results in 20 times more effective attack surface coverage than traditional methods.
All of this is important because skilled security staff are in short supply. 3.5 million cybersecurity positions are expected to go unfilled by 2021. Only 14 percent of IT Managers believe they have the cyber skills they need on staff, and analysts are spending up to 15 minutes every hour on triaging and reviewing false positives.
AI can help by taking on repetitive tasks to find the most common types of cyber threat, and by conducting reconnaissance to build a more in-depth threat landscape. Plus it can be used for cybersecurity data analysis where AI can complete tasks with consistently higher accuracy than human analysts.
"While humans can't scale, machines can't think," says Dr Mark Kuhr, CTO and co-founder of Synack. "More than 70 percent of the vulnerabilities that our Synack Red Team find in digital assets aren't detected by a traditional scanner. We will always need the creativity of human intelligence. But to scale at the pace of the threats, we need to keep building augmenting technology to test 'smarter'."
By combining human intelligence and artificial intelligence, companies are able to find and close critical security vulnerabilities 40 percent faster. Using this augmented approach also delivers four times greater return on investment than purely human testing methods.
You can get the full Trust at Scale report from the Synack website.